Upload 2 files

app.py

@@ -0,0 +1,717 @@
+"""
+╔══════════════════════════════════════════════════════════╗
+║         EMAIL JOB TRACKER — Hugging Face Spaces          ║
+║         SDK: Gradio  |  Model: LLaMA-3 via Groq          ║
+╚══════════════════════════════════════════════════════════╝
+
+Deploy:
+  1. Create a new HF Space → SDK: Gradio
+  2. Upload app.py + requirements.txt
+  3. Space auto-builds and launches
+"""
+
+import gradio as gr
+import imapclient
+import pyzmail
+import pandas as pd
+import json
+import os
+from datetime import datetime
+from openai import OpenAI
+
+# ──────────────────────────────────────────────
+# CONSTANTS
+# ──────────────────────────────────────────────
+IMAP_SERVER = "imap.gmail.com"
+LEARN_FILE  = "/tmp/learned_keywords.json"
+
+GMAIL_FOLDERS = [
+    "INBOX",
+    "[Gmail]/Sent Mail",
+    "[Gmail]/Drafts",
+    "[Gmail]/Important",
+    "[Gmail]/Starred",
+]
+
+CATEGORY_COLORS = {
+    "Offer":           "#10b981",
+    "Interview":       "#3b82f6",
+    "New Opportunity": "#8b5cf6",
+    "Rejected":        "#f43f5e",
+    "Urgent":          "#f97316",
+    "Alerts":          "#facc15",
+    "Spam":            "#6b7280",
+    "Unknown":         "#94a3b8",
+}
+
+CATEGORY_ICONS = {
+    "Offer":           "🎉",
+    "Interview":       "📅",
+    "New Opportunity": "🚀",
+    "Rejected":        "❌",
+    "Urgent":          "🔥",
+    "Alerts":          "🔔",
+    "Spam":            "🗑️",
+    "Unknown":         "❓",
+}
+
+# ──────────────────────────────────────────────
+# SELF-LEARNING KEYWORDS
+# ──────────────────────────────────────────────
+def load_learnings():
+    try:
+        if os.path.exists(LEARN_FILE):
+            with open(LEARN_FILE) as f:
+                return json.load(f)
+    except Exception:
+        pass
+    return {}
+
+def save_learnings(lk):
+    try:
+        with open(LEARN_FILE, "w") as f:
+            json.dump(lk, f, indent=2)
+    except Exception:
+        pass
+
+def get_base_keywords():
+    return {
+        "Offer": [
+            "offer letter", "congratulations", "selected", "employment offer",
+            "welcome aboard", "joining", "ctc", "salary", "compensation",
+            "offer rollout", "final selection", "offer acceptance",
+        ],
+        "Interview": [
+            "interview", "round", "technical", "hr round", "final round",
+            "assessment", "test", "assignment", "panel", "discussion",
+            "screening", "shortlisted", "interview invite",
+        ],
+        "New Opportunity": [
+            "job opportunity", "hiring", "opening", "vacancy", "role opportunity",
+            "we are hiring", "position available", "jd attached",
+            "job description", "career opportunity", "looking for candidates",
+        ],
+        "Rejected": [
+            "regret", "unfortunately", "not selected", "not shortlisted",
+            "rejected", "not a fit", "position filled", "application unsuccessful",
+        ],
+        "Urgent": [
+            "urgent", "immediate", "asap", "priority", "today", "tomorrow",
+            "deadline", "action required", "important", "quick response",
+        ],
+        "Alerts": [
+            "alert", "notification", "reminder", "update", "system alert",
+            "security alert", "account alert", "important update",
+        ],
+        "Spam": [
+            "sale", "discount", "offer deal", "win", "lottery", "free",
+            "click here", "subscribe", "buy now", "limited offer",
+        ],
+    }
+
+def rule_based_classification(text, lk):
+    text = text.lower()
+    kw = get_base_keywords()
+    for cat in lk:
+        kw.setdefault(cat, []).extend(lk[cat])
+    scores = {c: sum(1 for w in words if w in text) for c, words in kw.items()}
+    best = max(scores, key=scores.get)
+    return best if scores[best] > 0 else "Unknown"
+
+def learn_from_email(text, category, lk):
+    filtered = [w for w in text.lower().split() if len(w) > 6]
+    lk.setdefault(category, [])
+    for w in filtered[:5]:
+        if w not in lk[category]:
+            lk[category].append(w)
+    save_learnings(lk)
+
+# ──────────────────────────────────────────────
+# LLM CLASSIFIER  (Groq / LLaMA-3)
+# ──────────────────────────────────────────────
+def classify_email(subject, body, client, lk):
+    text = (subject + " " + body)[:1500]
+    prompt = f"""Classify this email into exactly ONE of:
+New Opportunity, Interview, Spam, Unknown, Offer, Rejected, Urgent, Alerts
+
+Extract company name, role/position, and interview round if present.
+
+Email:
+{text}
+
+Return ONLY a valid JSON object — no markdown, no backticks:
+{{
+    "category": "",
+    "company": "",
+    "role": "",
+    "round": "",
+    "confidence": 0.0
+}}"""
+    try:
+        res = client.chat.completions.create(
+            model="llama3-70b-8192",
+            messages=[{"role": "user", "content": prompt}],
+            temperature=0,
+        )
+        output = res.choices[0].message.content.strip()
+        output = output.replace("```json", "").replace("```", "").strip()
+        result = json.loads(output)
+        if result.get("confidence", 0) > 0.8:
+            learn_from_email(text, result["category"], lk)
+        return result
+    except Exception:
+        cat = rule_based_classification(text, lk)
+        learn_from_email(text, cat, lk)
+        return {"category": cat, "company": "Unknown",
+                "role": "Unknown", "round": "N/A", "confidence": 0.6}
+
+# ──────────────────────────────────────────────
+# EMAIL FETCH
+# ──────────────────────────────────────────────
+def fetch_all_emails(email, password, limit):
+    mail = imapclient.IMAPClient(IMAP_SERVER, ssl=True)
+    mail.login(email, password)
+    collected = []
+    for folder in GMAIL_FOLDERS:
+        try:
+            mail.select_folder(folder, readonly=True)
+            uids = mail.search(["ALL"])[-limit:]
+            for uid in uids:
+                raw = mail.fetch(uid, ["BODY[]"])
+                msg = pyzmail.PyzMessage.factory(raw[uid][b"BODY[]"])
+                subj = msg.get_subject() or "(no subject)"
+                if msg.text_part:
+                    body = msg.text_part.get_payload().decode(
+                        msg.text_part.charset or "utf-8", errors="replace")
+                elif msg.html_part:
+                    body = msg.html_part.get_payload().decode(
+                        msg.html_part.charset or "utf-8", errors="replace")
+                else:
+                    body = ""
+                collected.append({"folder": folder, "subject": subj, "body": body})
+        except Exception:
+            continue
+    mail.logout()
+    return collected
+
+# ──────────────────────────────────────────────
+# HTML HELPERS
+# ──────────────────────────────────────────────
+def make_summary_cards(counts):
+    cards = ""
+    for cat, cnt in counts.items():
+        color = CATEGORY_COLORS.get(cat, "#94a3b8")
+        icon  = CATEGORY_ICONS.get(cat, "•")
+        cards += f"""
+        <div style="
+            background: linear-gradient(145deg, #0d1117, #161b22);
+            border: 1px solid {color}55;
+            border-left: 3px solid {color};
+            border-radius: 10px;
+            padding: 14px 20px;
+            min-width: 110px;
+            text-align: center;
+            box-shadow: 0 4px 20px {color}22;
+        ">
+            <div style="font-size:1.1rem; margin-bottom:4px;">{icon}</div>
+            <div style="font-size:2rem; font-weight:800; color:{color}; line-height:1; font-family:'Courier New',monospace;">{cnt}</div>
+            <div style="font-size:0.68rem; color:#8b949e; margin-top:5px; letter-spacing:.08em; text-transform:uppercase;">{cat}</div>
+        </div>"""
+    return f"""
+    <div style="display:flex; flex-wrap:wrap; gap:12px; padding:8px 0;">
+        {cards}
+    </div>"""
+
+def make_log_html(log_items):
+    rows = ""
+    for item in log_items:
+        color = CATEGORY_COLORS.get(item["category"], "#94a3b8")
+        icon  = CATEGORY_ICONS.get(item["category"], "•")
+        conf_pct = int(item["confidence"] * 100)
+        conf_color = "#10b981" if conf_pct >= 80 else "#facc15" if conf_pct >= 60 else "#f43f5e"
+        rows += f"""
+        <div style="
+            display:flex; align-items:center; gap:12px;
+            padding:10px 14px; border-bottom:1px solid #21262d;
+            transition: background .15s;
+        " onmouseover="this.style.background='#161b22'" onmouseout="this.style.background='transparent'">
+            <span style="
+                background:{color}22; color:{color};
+                border:1px solid {color}55;
+                padding:3px 10px; border-radius:20px;
+                font-size:.7rem; font-weight:700;
+                white-space:nowrap; min-width:90px; text-align:center;
+            ">{icon} {item["category"]}</span>
+            <span style="color:#8b949e; font-size:.72rem; white-space:nowrap; font-family:'Courier New',monospace;">[{item["folder"].replace("[Gmail]/","")}]</span>
+            <span style="color:#c9d1d9; font-size:.82rem; flex:1; overflow:hidden; text-overflow:ellipsis; white-space:nowrap;">{item["subject"][:75]}</span>
+            <span style="color:#58a6ff; font-size:.75rem; white-space:nowrap;">{item["company"]}</span>
+            <span style="color:{conf_color}; font-size:.7rem; font-family:'Courier New',monospace; white-space:nowrap;">{conf_pct}%</span>
+        </div>"""
+    return f"""
+    <div style="
+        background:#0d1117;
+        border:1px solid #21262d;
+        border-radius:10px;
+        overflow:hidden;
+        font-family: 'Courier New', monospace;
+    ">
+        <div style="
+            background:#161b22;
+            padding:10px 14px;
+            display:flex; gap:20px;
+            font-size:.68rem; color:#8b949e;
+            letter-spacing:.1em; text-transform:uppercase;
+            border-bottom:1px solid #21262d;
+        ">
+            <span style="min-width:110px;">Category</span>
+            <span style="min-width:70px;">Folder</span>
+            <span style="flex:1;">Subject</span>
+            <span style="min-width:80px;">Company</span>
+            <span>Conf.</span>
+        </div>
+        <div style="max-height:380px; overflow-y:auto;">
+            {rows}
+        </div>
+    </div>"""
+
+def make_pipeline_html(tracker_df):
+    """Group by category, show company cards."""
+    if tracker_df is None or tracker_df.empty:
+        return ""
+    sections = ""
+    for cat in tracker_df["Category"].unique():
+        color = CATEGORY_COLORS.get(cat, "#94a3b8")
+        icon  = CATEGORY_ICONS.get(cat, "•")
+        subset = tracker_df[tracker_df["Category"] == cat]
+        items = ""
+        for _, row in subset.iterrows():
+            items += f"""
+            <div style="
+                background:#161b22;
+                border:1px solid #30363d;
+                border-radius:8px;
+                padding:10px 14px;
+                margin-bottom:6px;
+            ">
+                <div style="font-size:.85rem; font-weight:700; color:#e6edf3;">{row.get("Company","—")}</div>
+                <div style="font-size:.75rem; color:#8b949e; margin-top:3px;">
+                    {row.get("Role","—")}
+                    {"&nbsp;·&nbsp;<span style='color:#58a6ff;'>"+row.get("Round","")+"</span>" if row.get("Round","N/A") not in ["N/A","","None"] else ""}
+                </div>
+            </div>"""
+        sections += f"""
+        <div style="margin-bottom:18px;">
+            <div style="
+                font-size:.7rem; letter-spacing:.15em; text-transform:uppercase;
+                color:{color}; margin-bottom:8px;
+                font-family:'Courier New',monospace;
+            ">{icon} {cat} ({len(subset)})</div>
+            {items}
+        </div>"""
+    return f"""
+    <div style="
+        background:#0d1117;
+        border:1px solid #21262d;
+        border-radius:10px;
+        padding:18px;
+        max-height:460px;
+        overflow-y:auto;
+    ">
+        {sections}
+    </div>"""
+
+# ──────────────────────────────────────────────
+# MAIN PIPELINE
+# ──────────────────────────────────────────────
+def run_pipeline(email, password, groq_key, limit, progress=gr.Progress(track_tqdm=True)):
+    if not email.strip() or not password.strip() or not groq_key.strip():
+        yield (
+            gr.update(value="<div style='color:#f43f5e;padding:16px;font-family:monospace;'>⚠️ Please fill in all three credential fields.</div>"),
+            gr.update(value=""), gr.update(value=""), None, None,
+        )
+        return
+
+    client = OpenAI(api_key=groq_key.strip(), base_url="https://api.groq.com/openai/v1")
+    lk     = load_learnings()
+
+    yield (
+        gr.update(value="<div style='color:#facc15;padding:16px;font-family:monospace;'>🔌 Connecting to Gmail…</div>"),
+        gr.update(value=""), gr.update(value=""), None, None,
+    )
+
+    try:
+        emails = fetch_all_emails(email.strip(), password.strip(), int(limit))
+    except Exception as e:
+        yield (
+            gr.update(value=f"<div style='color:#f43f5e;padding:16px;font-family:monospace;'>❌ Gmail failed: {e}<br><br>Check email address and App Password.</div>"),
+            gr.update(value=""), gr.update(value=""), None, None,
+        )
+        return
+
+    if not emails:
+        yield (
+            gr.update(value="<div style='color:#94a3b8;padding:16px;font-family:monospace;'>No emails found in selected folders.</div>"),
+            gr.update(value=""), gr.update(value=""), None, None,
+        )
+        return
+
+    records, log_items = [], []
+    total = len(emails)
+
+    for i, mail in enumerate(emails):
+        progress((i + 1) / total)
+        result = classify_email(mail["subject"], mail["body"], client, lk)
+        rec = {
+            "Timestamp":  datetime.now().strftime("%Y-%m-%d %H:%M"),
+            "Folder":     mail["folder"],
+            "Company":    result.get("company",    "Unknown"),
+            "Role":       result.get("role",       "Unknown"),
+            "Category":   result.get("category",   "Unknown"),
+            "Round":      result.get("round",      "N/A"),
+            "Confidence": round(result.get("confidence", 0.5), 2),
+            "Subject":    mail["subject"],
+        }
+        records.append(rec)
+        log_items.append({
+            "folder":     mail["folder"],
+            "subject":    mail["subject"],
+            "category":   rec["Category"],
+            "company":    rec["Company"],
+            "confidence": rec["Confidence"],
+        })
+
+        # Live update every 5 emails
+        if (i + 1) % 5 == 0 or (i + 1) == total:
+            df_partial  = pd.DataFrame(records)
+            partial_log = make_log_html(log_items)
+            partial_sum = make_summary_cards(df_partial["Category"].value_counts())
+            yield (
+                gr.update(value=partial_sum),
+                gr.update(value=partial_log),
+                gr.update(value=""),
+                None, None,
+            )
+
+    df      = pd.DataFrame(records)
+    tracker = df.groupby(["Company", "Role"]).last().reset_index()
+    df.to_csv("/tmp/email_log.csv",    index=False)
+    tracker.to_csv("/tmp/job_tracker.csv", index=False)
+
+    summary_html  = make_summary_cards(tracker["Category"].value_counts())
+    log_html      = make_log_html(log_items)
+    pipeline_html = make_pipeline_html(tracker)
+
+    yield (
+        gr.update(value=summary_html),
+        gr.update(value=log_html),
+        gr.update(value=pipeline_html),
+        "/tmp/email_log.csv",
+        "/tmp/job_tracker.csv",
+    )
+
+# ──────────────────────────────────────────────
+# GRADIO UI
+# ──────────────────────────────────────────────
+CSS = """
+@import url('https://fonts.googleapis.com/css2?family=JetBrains+Mono:wght@300;400;600;800&family=DM+Sans:wght@300;400;500;700&display=swap');
+
+*, *::before, *::after { box-sizing: border-box; }
+
+body, .gradio-container {
+    background: #010409 !important;
+    color: #e6edf3 !important;
+    font-family: 'DM Sans', sans-serif !important;
+}
+
+/* Tabs */
+.tab-nav { background: #0d1117 !important; border-bottom: 1px solid #21262d !important; }
+.tab-nav button { color: #8b949e !important; font-family: 'DM Sans', sans-serif !important; font-size:.85rem !important; }
+.tab-nav button.selected { color: #58a6ff !important; border-bottom: 2px solid #58a6ff !important; background: transparent !important; }
+
+/* Inputs */
+label > span {
+    color: #8b949e !important;
+    font-size: .7rem !important;
+    letter-spacing: .12em !important;
+    text-transform: uppercase !important;
+    font-family: 'JetBrains Mono', monospace !important;
+}
+input[type=text], input[type=password], textarea {
+    background: #0d1117 !important;
+    border: 1px solid #30363d !important;
+    color: #e6edf3 !important;
+    border-radius: 6px !important;
+    font-family: 'DM Sans', sans-serif !important;
+    font-size: .88rem !important;
+    transition: border-color .2s !important;
+}
+input[type=text]:focus, input[type=password]:focus {
+    border-color: #58a6ff !important;
+    outline: none !important;
+    box-shadow: 0 0 0 3px #58a6ff18 !important;
+}
+
+/* Range slider */
+input[type=range] { accent-color: #58a6ff !important; }
+
+/* Buttons */
+.gr-button {
+    font-family: 'DM Sans', sans-serif !important;
+    border-radius: 6px !important;
+    font-weight: 600 !important;
+    transition: all .2s !important;
+}
+.gr-button-primary {
+    background: #238636 !important;
+    border: 1px solid #2ea043 !important;
+    color: #ffffff !important;
+}
+.gr-button-primary:hover {
+    background: #2ea043 !important;
+    transform: translateY(-1px) !important;
+    box-shadow: 0 4px 16px #2ea04344 !important;
+}
+.gr-button-secondary {
+    background: #21262d !important;
+    border: 1px solid #30363d !important;
+    color: #c9d1d9 !important;
+}
+
+/* Panels */
+.gr-panel, .gr-box, .gr-form {
+    background: #0d1117 !important;
+    border: 1px solid #21262d !important;
+    border-radius: 10px !important;
+}
+
+/* Dataframe */
+.gr-dataframe table {
+    background: #0d1117 !important;
+    border: 1px solid #21262d !important;
+    font-family: 'JetBrains Mono', monospace !important;
+    font-size: .78rem !important;
+}
+.gr-dataframe th {
+    background: #161b22 !important;
+    color: #8b949e !important;
+    letter-spacing: .06em !important;
+    text-transform: uppercase !important;
+}
+.gr-dataframe td { color: #c9d1d9 !important; border-color: #21262d !important; }
+.gr-dataframe tr:hover td { background: #161b22 !important; }
+
+/* Scrollbar */
+::-webkit-scrollbar { width: 6px; height: 6px; }
+::-webkit-scrollbar-track { background: #0d1117; }
+::-webkit-scrollbar-thumb { background: #30363d; border-radius: 3px; }
+::-webkit-scrollbar-thumb:hover { background: #484f58; }
+"""
+
+HEADER_HTML = """
+<div style="
+    padding: 32px 0 24px;
+    font-family: 'DM Sans', sans-serif;
+    border-bottom: 1px solid #21262d;
+    margin-bottom: 24px;
+">
+    <div style="display:flex; align-items:center; gap:16px; flex-wrap:wrap;">
+        <div style="
+            background: linear-gradient(135deg, #238636, #3fb950);
+            border-radius: 12px;
+            padding: 12px 16px;
+            font-size: 1.6rem;
+            line-height:1;
+        ">📬</div>
+        <div>
+            <div style="
+                font-size:.65rem; letter-spacing:.25em; text-transform:uppercase;
+                color:#58a6ff; font-family:'JetBrains Mono',monospace; margin-bottom:4px;
+            ">Powered by LLaMA-3 · Self-Learning · Groq</div>
+            <h1 style="
+                font-size:1.9rem; font-weight:800; color:#e6edf3;
+                margin:0; letter-spacing:-.03em; line-height:1;
+            ">Email Job Tracker</h1>
+            <p style="color:#8b949e; margin:6px 0 0; font-size:.88rem; font-weight:300;">
+                Fetch → classify → track your entire job pipeline automatically
+            </p>
+        </div>
+        <div style="margin-left:auto; text-align:right; display:flex; flex-direction:column; gap:6px;">
+            <span style="
+                background:#238636; color:#fff; font-size:.68rem;
+                padding:3px 10px; border-radius:20px; font-weight:600;
+                font-family:'JetBrains Mono',monospace; letter-spacing:.05em;
+            ">v2.0 GRADIO</span>
+            <span style="
+                background:#1f6feb22; color:#58a6ff; font-size:.68rem;
+                padding:3px 10px; border-radius:20px;
+                font-family:'JetBrains Mono',monospace; border:1px solid #1f6feb55;
+            ">HF SPACES</span>
+        </div>
+    </div>
+</div>
+"""
+
+CRED_HELP = """
+<div style="
+    background: #161b22;
+    border: 1px solid #30363d;
+    border-left: 3px solid #58a6ff;
+    border-radius: 8px;
+    padding: 14px 16px;
+    font-size: .78rem;
+    color: #8b949e;
+    font-family: 'DM Sans', sans-serif;
+    line-height: 1.7;
+">
+    <strong style="color:#c9d1d9;">🔐 Security Notes</strong><br>
+    • Use a <strong style="color:#e6edf3;">Gmail App Password</strong>, not your account password<br>
+    • Requires 2FA enabled →
+      <a href="https://myaccount.google.com/apppasswords" target="_blank" style="color:#58a6ff;">Generate here</a><br>
+    • Get free Groq API key →
+      <a href="https://console.groq.com" target="_blank" style="color:#58a6ff;">console.groq.com</a><br>
+    • Credentials are <strong style="color:#e6edf3;">never stored</strong> — session only
+</div>
+"""
+
+EMPTY_LOG = """
+<div style="
+    background:#0d1117; border:1px solid #21262d; border-radius:10px;
+    padding:40px; text-align:center; color:#484f58;
+    font-family:'JetBrains Mono',monospace; font-size:.82rem;
+">
+    ░░ No data yet — run the agent to see results ░░
+</div>"""
+
+def section_label(text, color="#58a6ff"):
+    return f"""<div style="
+        font-size:.65rem; letter-spacing:.18em; text-transform:uppercase;
+        color:{color}; font-family:'JetBrains Mono',monospace;
+        margin-bottom:8px; margin-top:4px;
+    ">{text}</div>"""
+
+with gr.Blocks(css=CSS, title="📬 Email Job Tracker") as demo:
+    gr.HTML(HEADER_HTML)
+
+    with gr.Tabs():
+
+        # ════════════════════════════════════════
+        # TAB 1 — RUN AGENT
+        # ════════════════════════════════════════
+        with gr.TabItem("⚡ Run Agent"):
+            with gr.Row(equal_height=False):
+
+                # LEFT — credentials
+                with gr.Column(scale=1, min_width=300):
+                    gr.HTML(section_label("Credentials"))
+                    email_in   = gr.Textbox(label="Gmail Address",      placeholder="you@gmail.com")
+                    pass_in    = gr.Textbox(label="Gmail App Password",  type="password", placeholder="xxxx xxxx xxxx xxxx")
+                    groq_in    = gr.Textbox(label="Groq API Key",        type="password", placeholder="gsk_...")
+                    gr.HTML(section_label("Settings", "#8b949e"))
+                    limit_in   = gr.Slider(label="Emails per folder", minimum=5, maximum=50, value=20, step=5)
+                    run_btn    = gr.Button("⚡  Run Agent", variant="primary", size="lg")
+                    clear_btn  = gr.Button("✕  Clear Results", variant="secondary", size="sm")
+                    gr.HTML(CRED_HELP)
+
+                # RIGHT — live results
+                with gr.Column(scale=2):
+                    gr.HTML(section_label("Category Summary"))
+                    summary_out = gr.HTML(value=EMPTY_LOG)
+
+                    gr.HTML(section_label("Live Classification Log"))
+                    log_out = gr.HTML(value=EMPTY_LOG)
+
+            gr.HTML(section_label("Job Pipeline View"))
+            pipeline_out = gr.HTML(value=EMPTY_LOG)
+
+            with gr.Row():
+                with gr.Column():
+                    gr.HTML(section_label("Downloads", "#8b949e"))
+                    file_log     = gr.File(label="📄 email_log.csv — all classified emails")
+                    file_tracker = gr.File(label="📊 job_tracker.csv — deduplicated pipeline")
+
+        # ════════════════════════════════════════
+        # TAB 2 — TRACKER TABLE
+        # ════════════════════════════════════════
+        with gr.TabItem("📊 Tracker Table"):
+            gr.HTML(section_label("Full Job Tracker — Latest status per company/role"))
+            table_out = gr.Dataframe(
+                headers=["Company", "Role", "Category", "Round", "Confidence", "Subject"],
+                interactive=False,
+                wrap=False,
+            )
+
+        # ════════════════════════════════════════
+        # TAB 3 — HOW TO USE
+        # ════════════════════════════════════════
+        with gr.TabItem("📖 Guide"):
+            gr.HTML("""
+<div style="max-width:680px; font-family:'DM Sans',sans-serif; line-height:1.8; padding:8px 0;">
+
+<h2 style="color:#e6edf3; font-size:1.1rem; margin-bottom:4px;">🚀 Quick Start</h2>
+<ol style="color:#8b949e; font-size:.88rem; padding-left:20px;">
+  <li>Enable 2-Step Verification on your Google Account</li>
+  <li>Generate a <a href="https://myaccount.google.com/apppasswords" target="_blank" style="color:#58a6ff;">Gmail App Password</a> (16 chars, no spaces)</li>
+  <li>Get a free API key from <a href="https://console.groq.com" target="_blank" style="color:#58a6ff;">Groq Console</a></li>
+  <li>Enter credentials in the <strong style="color:#c9d1d9;">Run Agent</strong> tab and hit ⚡ Run</li>
+</ol>
+
+<h2 style="color:#e6edf3; font-size:1.1rem; margin-top:20px; margin-bottom:4px;">🏷️ Email Categories</h2>
+<table style="font-size:.82rem; border-collapse:collapse; width:100%;">
+  <thead>
+    <tr style="background:#161b22;">
+      <th style="padding:8px 12px; color:#8b949e; text-align:left; border:1px solid #21262d;">Category</th>
+      <th style="padding:8px 12px; color:#8b949e; text-align:left; border:1px solid #21262d;">Triggers</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr><td style="padding:8px 12px; color:#10b981; border:1px solid #21262d;">🎉 Offer</td><td style="padding:8px 12px; color:#8b949e; border:1px solid #21262d;">Offer letter, CTC, Welcome aboard</td></tr>
+    <tr><td style="padding:8px 12px; color:#3b82f6; border:1px solid #21262d;">📅 Interview</td><td style="padding:8px 12px; color:#8b949e; border:1px solid #21262d;">Invite, Round, Assessment, Screening</td></tr>
+    <tr><td style="padding:8px 12px; color:#8b5cf6; border:1px solid #21262d;">🚀 New Opportunity</td><td style="padding:8px 12px; color:#8b949e; border:1px solid #21262d;">Hiring, Opening, JD Attached</td></tr>
+    <tr><td style="padding:8px 12px; color:#f43f5e; border:1px solid #21262d;">❌ Rejected</td><td style="padding:8px 12px; color:#8b949e; border:1px solid #21262d;">Regret, Unfortunately, Not selected</td></tr>
+    <tr><td style="padding:8px 12px; color:#f97316; border:1px solid #21262d;">🔥 Urgent</td><td style="padding:8px 12px; color:#8b949e; border:1px solid #21262d;">ASAP, Deadline, Action Required</td></tr>
+    <tr><td style="padding:8px 12px; color:#facc15; border:1px solid #21262d;">🔔 Alerts</td><td style="padding:8px 12px; color:#8b949e; border:1px solid #21262d;">Security Alert, Account Notification</td></tr>
+    <tr><td style="padding:8px 12px; color:#6b7280; border:1px solid #21262d;">🗑️ Spam</td><td style="padding:8px 12px; color:#8b949e; border:1px solid #21262d;">Sale, Lottery, Buy Now, Subscribe</td></tr>
+  </tbody>
+</table>
+
+<h2 style="color:#e6edf3; font-size:1.1rem; margin-top:20px; margin-bottom:4px;">🧠 Self-Learning</h2>
+<p style="color:#8b949e; font-size:.88rem;">
+Every high-confidence classification (>80%) teaches the system new keywords for that category. These are saved to <code style="color:#58a6ff;">/tmp/learned_keywords.json</code> and improve accuracy over time within the session.
+</p>
+
+</div>
+""")
+
+    # ════════════════════════════════════════
+    # EVENT HANDLERS
+    # ════════════════════════════════════════
+    def do_run(email, password, groq_key, limit, progress=gr.Progress(track_tqdm=True)):
+        tracker_df_holder = []
+        last_yield = None
+        for result in run_pipeline(email, password, groq_key, limit, progress):
+            last_yield = result
+            yield result[0], result[1], result[2], result[3], result[4], gr.update()
+
+        # populate tracker table after run
+        try:
+            df = pd.read_csv("/tmp/job_tracker.csv")
+            cols = [c for c in ["Company","Role","Category","Round","Confidence","Subject"] if c in df.columns]
+            yield last_yield[0], last_yield[1], last_yield[2], last_yield[3], last_yield[4], df[cols]
+        except Exception:
+            yield last_yield[0], last_yield[1], last_yield[2], last_yield[3], last_yield[4], gr.update()
+
+    def do_clear():
+        return EMPTY_LOG, EMPTY_LOG, EMPTY_LOG, None, None, None
+
+    run_btn.click(
+        fn=do_run,
+        inputs=[email_in, pass_in, groq_in, limit_in],
+        outputs=[summary_out, log_out, pipeline_out, file_log, file_tracker, table_out],
+    )
+    clear_btn.click(
+        fn=do_clear,
+        outputs=[summary_out, log_out, pipeline_out, file_log, file_tracker, table_out],
+    )
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,5 @@
+gradio>=4.0.0
+imapclient>=2.3.1
+pyzmail36
+openai>=1.0.0
+pandas>=2.0.0